Analytics BIOC Low

Suspicious AI Dataset Label Modification

AI Dataset labels were modified by an identity that typically doesn't interact with labels. MITRE ATLAS Technique: AML.T0020 - Poison Training Data. OWASP Top 10 LLM Technique: LLM04 - Data and Model Poisoning.

Module:
Cortex Cloud
Licensed by:
Cloud Runtime Security (CRS)
Data source:
AWS Audit Log, Azure Audit Log, Gcp Audit Log
ATT&CK tactics: Impact (TA0040)
ATT&CK techniques: Data Manipulation: Stored Data Manipulation (T1565.001)
Detector tags: Cloud AI Infrastructure Analytics
Attacker's goals:

Contaminating training set, so that predictions on new data will be modified.

Investigative actions:

Check the identity that modified the dataset's labels. Check that the dataset is correctly labeled.

Test period:
N/A (single event)
Deduplication:
5 Days